[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <C595599E-0BF9-462E-A79B-E7F6C9D5F587@zytor.com>
Date: Fri, 03 Oct 2025 09:53:24 -0700
From: "H. Peter Anvin" <hpa@...or.com>
To: Dave Hansen <dave.hansen@...el.com>, Peter Zijlstra <peterz@...radead.org>,
Jürgen Groß <jgross@...e.com>
CC: linux-kernel@...r.kernel.org, x86@...nel.org, llvm@...ts.linux.dev,
xin@...or.com, Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
Dave Hansen <dave.hansen@...ux.intel.com>,
Nathan Chancellor <nathan@...nel.org>,
Nick Desaulniers <nick.desaulniers+lkml@...il.com>,
Bill Wendling <morbo@...gle.com>,
Justin Stitt <justinstitt@...gle.com>
Subject: Re: [PATCH v2 09/12] x86/msr: Use the alternatives mechanism for WRMSR
On October 3, 2025 7:23:29 AM PDT, Dave Hansen <dave.hansen@...el.com> wrote:
>On 10/1/25 00:23, Peter Zijlstra wrote:
>> On Wed, Oct 01, 2025 at 08:43:39AM +0200, Peter Zijlstra wrote:
>>> Let me see how terrible it all ends up when using as macros
>> Argh, as macros are differently painful. I hate computers :/
>
>ALTERNATIVES are fun and all, but is there a good reason we're pulling
>out our hair to use them here?
>
>Normal WRMSR is slooooooooooow. The ones that aren't slow don't need
>WRMSRNS in the first place.
>
>Would an out-of-line wrmsr() with an if() in it be so bad? Or a
>static_call()? Having WRMSR be inlined in a laudable goal, but I'm
>really asking if it's worth it.
We need them to use wrmsrns immediate.
Powered by blists - more mailing lists